Materials Chemistry and Physics 136 (2012) 394e414

Contents lists available at SciVerse ScienceDirect

Materials Chemistry and Physics

journal homepage: www.elsevier.com/locate/matchemphys

Corrosion and wear resistance of hypoeutectic ZneAl alloys as a function

of structural features
A.E. Ares a, b, *, L.M. Gassa a, c, C.E. Schvezov a, b, M.R. Rosenberger a, b
a
Career of Scientific Research (CIC) of the National Science Research Council of Argentina (CONICET), Argentina
b
Faculty of Sciences, University of Misiones, 1552 Félix de Azara Street, 3300 Posadas-Misiones, Argentina
c
The Research Institute of Theoretical and Applied Physical Chemistry (INIFTA), National University of La Plata-CONICET, Suc 4, C.C. 16, (1900) La Plata, Argentina

h i g h l i g h t s

< A combined study encompassing the grain and dendritic scale was here performed.
< The results include corrosion as well as wear parameters.
< Even at the highest Al wt%, the corrosion resistance depends on the structure.
< The wear resistance increases from the columnar to the equiaxed structure.
< A clear inverse relation is observed between the wear resistance and l2.

a r t i c l e i n f o a b s t r a c t

Article history: The aim of the present research was to investigate the role of the type of grain macrostructure (columnar,
Received 2 September 2011 columnar-to-equiaxed transitioneCET -, and equiaxed) and microstructure (secondary dendrite arm
Received in revised form spacing - l2) in the wear and electrochemical behavior of hypereutectic ZneAl alloys (Zn-1wt%Al, Zn-2wt
2 June 2012
%Al, Zn-3wt%Al and Zn-4wt%Al) directionally solidified in a vertical upward directional solidification
Accepted 29 June 2012
device. We also aimed to, correlate the thermal parameters with the electrochemical and wear prop-
erties. The results obtained in this alloy system indicate that the corrosion resistance decreases as the
Keywords:
concentration of aluminum increases, but that at the highest concentrations (near the eutectic
Alloys
Solidification
concentration) this resistance depends on the structure. For the same wear conditions, the wear rate of
Tribology and wear the equiaxed region is lower than the columnar and transition regions. Independently of the type of
Corrosion test structure, wear resistance increases as the aluminum concentration increases.
Microstructure For each alloy concentration, the wear resistance increases from the columnar to the equiaxed
Corrosion structure. A clear inverse relation is observed between the wear resistance and l2, which may explain the
Mechanical properties increase in wear resistance in the equiaxed region. At low spacings, the wear rate converges to a single
value.
Ó 2012 Elsevier B.V. All rights reserved.

1. Introduction solidification rate is determined mainly by heat extraction through

Most manufacturing processes involve melting and solidifica-
tion of metals and alloys during the fabrication of various compo-
nents. The thermal and solutal conditions that prevail during the
manufacturing process and the thermodynamic and kinetic
constraints of the material determine the final microstructure of
the component. Solidification involves the extraction of heat from
the liquid and the motion of the solideliquid interface. The
* Corresponding author. Faculty of Sciences, University of Misiones, 1552 Félix de
Azara Street, 3300 Posadas-Misiones, Argentina. Tel.: þ54 3752 422186x156;
fax: þ54 3752 425414.
E-mail addresses: aares@fceqyn.unam.edu.ar, a.e.ares@gmail.com (A.E. Ares).
thermal diffusion and convection. However, the solidification
microstructure is a complex function of the composition of the
alloy, the solidification rate, the heat extraction flow, the temper-
ature gradients, and several material characteristics such as phase
equilibrium reactions, nucleation and growth kinetics of the phases
and crystallographic constraints.
A typical cast structure consists of one or both types of grains:
columnar or equiaxed. When both are present, the columnar is the
first to solidify and then the equiaxed grains are usually formed in
the central part of the cast. In such case, the transition is defined as
columnar-to-equiaxed transition (CET). A low solidification rate
favors the CET phenomenon as well as the addition of nucleant
particles [1].

0254-0584/$ e see front matter Ó 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.matchemphys.2012.06.065
 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414
During many manufacturing processes, the CET is normally
avoided because the casting must usually consist of only one type of
grain; for instance, columnar in the case of turbine blades and
equiaxed in the case of car engine parts.
The macrostructure of cast alloys is determined by process
parameters such as the superheat, alloy system, composition,
casting size, fluid flow, addition of grain refiner, and mechanical
disturbance [2]. Although the directional solidification process is
controlled mainly by temperature gradients and an important
contribution of flow, the structure is mostly affected by the
combined effect [1,2].
However, the CET structure is frequently not required and it is
very important to understand the interaction of equiaxed grains
between themselves and with the columnar front. In addition, it is
important to understand the physical mechanisms which control
the phenomenon and to link the parameters characterizing the CET
during the directional solidification process with mechanical and
corrosion properties of each type of structure (columnar, CET and
equiaxed).
A number of experimental investigations have suggested that
the CET occurs when the temperature gradient in the melt reaches
a minimum critical value [1e13]. Gandin [14] proposed a CET
criterion based on the position of the maximum velocity of the
columnar/dendritic interface, suggesting a continuous increase in
tip growth rates up to a maximum value about two-thirds the
length of the ingot, where the transition is supposed to occur. We
have previously investigated experimentally the directional solid-
ification of AleZn and ZneAl alloys under different conditions of
superheat [15] and found that the results are related to the solidi-
fication thermal parameters and recalescence determined from the
temperature versus time curves. The observations indicate that the
CET is the result of a competition between coarse columnar
dendrites and finer equiaxed dendrites. The experiments carried
out by Nguyen-Thi et al. [16] on Al-3.5wt%Ni alloy provided direct
access to dynamical phenomena during columnar growth, and for
the first time to the CET microstructure. In those experiments, the
dendritic morphologies were analyzed as a function of the solidi-
fication parameters. Siqueira et al. [5,17] and Spinelli et al. [18]
proposed a CET criterion based on critical cooling rate of about
0.014 K s1 and 0.030 K s1, respectively, with the columnar growth
prevailing throughout the casting for cooling rates higher than
these critical values. Badillo and Beckermann [19] developed
a study to perform direct numerical simulations of the CET in two
spatial dimensions, of the CET in directional solidification of an
Ale3wt%Cu alloy using the phase-field model.
The results illustrate the full complexities of the solute
concentration fields and solidification morphologies. Canté et al.
[20] conducted experiments to examine the CET during the upward
vertical directional solidification of hypoeutectic AleNi and AleSn
alloys under unsteady-state heat flow conditions and noted that
the CET occurs essentially on a near horizontal plane in the castings
when critical cooling rate in the melt is reached. For the investi-
gated alloys the average critical values were found to be about
0.16 K s1 for the AleNi alloys and 0.30 K s1 for the AleSn alloys.
Other previous studies of the CET develop expressions, simu-
lations and numerical procedures to describe a criterion for the CET
[17,19,21e34].
It is well known that convection influences the CET and primary
cellular/dendritic spacings. Natural convection during solidification
is principally caused by density gradients in the liquid. The
temperature gradient in the liquid may lead to thermal convection
and the rejected solute in the interdendritic liquid may cause sol-
utal convection. Whether the thermal and solutal buoyancy forces
may either counteract or increase each other depending on the
relative densities of the alloy constituents and the particular
395
constituent rejected into the interdendritic liquid during solidifi-
cation [35,36].
The uses of Zn can be divided into six major categories: (a)
coatings, (b) casting alloys, (c) alloying element in brass and other
alloys, (d) wrought zinc alloys, (e) zinc oxide, and (f) zinc chemicals.
The binary zinc alloy systems of most interest for commercial
applications are ZneAl, which at 4wt.%Al form the basis of the zinc
die-casting alloys. Cast zinc products are produced mainly by die-
casting process, in which the liquid metal is forced under pres-
sure into a cooled die and solidifies almost instantaneously to
produce a fine-grained product. One unique application of zinc
takes particular advantage of its ability to transfer its corrosion
resistance properties by electrical contact. This application is called
a sacrificial anode. The anodes, made of almost pure zinc, are bolted
to aluminum marine engines. During operation in water, especially
salt water, the oxidation forms a weak electrical current, which may
corrode the hull and engine parts. Since zinc is easily oxidized in
the presence of this electrical current, it “sacrifices” itself by
corroding quickly, consuming the entire electrical imbalance in the
ship. As a result, the remaining aluminum hull and engine are not
corroded as a result. As it is consumed, the anode must be replaced
to assure continued protection. In an application similar to the
sacrificial anode, zinc is used as a component in battery production
[37].
One particular area of research has been dedicated to the rela-
tionship between the properties and the cast structure. Sahoo et al.
[38], Prasad et al. [39] and Sriram et al. [40] investigated the effects
of trace elements on the casting characteristics and mechanical
properties of cast ZneAl alloys, and found that ultimate tensile
strength, yield strength and percent elongation of the alloys are
slightly lower in Sr-treated melts, but higher when Be is added.
They also found that hardness values increase when Be is added,
that impact strength is drastically reduced when Sr and B are
added, and that the elongation and impact strength are improved
when Sr and B are added.
Osorio et al. [41] studied the role of macrostructural
morphology and grain size in the corrosion resistance of Zn and Al
castings and concluded that a better corrosion resistance tendency
is achieved with coarse macrostructures rather than with fine
grains for both Zn and Al. When studying the effect of the dendritic
microstructure of the corrosion resistance of ZneAl alloys Osorio
et al. [42] found that the tendency of improvement on the corrosion
resistance depends both on the cooling rate imposed during
solidification, which affects dendrite arm size and the solute
redistribution, and on the electrochemical behavior of solute and
solvent. In addition, by studying the application of electrochemical
impedance spectroscopy to investigate the effect of as-cast struc-
tures on the corrosion resistance of hypereutectic ZneAl alloys,
these authors showed that the corrosion resistance of hypereu-
tectic ZneAl alloys tends to improve as the dendritic arm spacing
increases and that the impedance parameters of both zones
(columnar and equiaxed) are similar. In their experiments, the CET
occurred in a sharp plane which did not yield a mixed structure and
therefore did not characterize the properties of the mixed structure
formed with columnar and equiaxed grains [43].
Abou El-khair et al. [44] studied the effects of different Al
contents on the microstructure, hardness, tensile properties and
wear behavior of Zn-based alloys and found that the hardness of
ZneAl alloys increases as Al content is increased from 8 to 27 wt.%,
and that there is a significant improvement in ultimate strength at
room and high temperatures.
Auras and Schvezov [45] studied the influence of alloying
elements and SiC particles in the solidification process, alloy
morphology, wear mechanism and debris formation in ZA27-based
alloys containing silicon and copper and found the following: that
396 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414

the Si precipitates appear in a plate-like shape in the dendrites or Table 1

grain, that the cooper appears in the solution at concentrations Chemical composition of the Zn and Al used to prepare the
alloys.
lower than 1 pct and as a precipitate in the interdendritic at higher
concentrations, and that the SiC particles appear in the interden- Element Weight percent, wt%
dritic regions. Chemical composition of Zn
Manna et al. [46e50] studied the formation of a precipitate Zn 99.98  0.2
Fe 0.010  0.01
phase b and a solute depleted matrix phase a behind a migrating
Si 0.006  0.0001
boundary advancing into a supersaturated matrix phase a0 which is Pb 0.004  0.001
called discontinuous precipitation (DP) [51] and found that the Others < 0.001  0.0001
precipitation is characterized by discontinuous discontinuous Chemical composition of Al
changes in orientation and composition between the matrix phases Al 99.94  0.2
Fe 0.028  0.0001
(a0/a) across the migrating boundary, called the reaction front (RF), Si 0.033  0.001
which provides a short-circuit path of solute transport [51e53]. Pb 0.001  0.0001
They also found that the Zn-2at.%Ag alloy undergoes a discontin- Others < 0.001  0.0001
uous mode of precipitation in the temperature range of 353e433 K.
The rod-like precipitate phase in the DP colony maintains
a distance of separation statistically constant under isothermal
conditions of precipitation. This approach has already been applied was independent of the structure, whereas the alloy with 27wt.%Al
in several binary systems like ZneAl [54] and other systems and CET structure was the most resistant of all. By correlating the
[55e58]. EIS parameters with thermal parameters like Rct with the critical
In previous works, we studied the influence of thermal param- temperature gradient, Gc, we found that Rct increases when Gc
eters on the structure and CET transition [13,15,32,59] and becomes more negative. In the case of the correlation between Rct
measured the mechanical and corrosion properties of the alloys and structural parameters, i.e. grain size and secondary dendrite
with different structures [60,61]. Our main results indicated that spacing, Rct also increased when both parameters increased. This
the CET occurs in a zone rather than in a sharp plane, where both was not observed in Zn-16wt.%Al alloys.
columnar and equiaxed grains coexist with the melt, and that the Another recent research shows that what actually affects the
length of the columnar zone increases with the cooling rate and response to corrosion is the way in which aluminum is distributed
decreases with alloy composition, which could be associated with in the alloy, i.e., which phases are present in the solidified micro-
the kinetics of solidification. Also, the CET was observed to occur structure and how they are distributed, and not the amount of
when the temperature gradient in the melt decreased to values aluminum present in the alloy [64].
close to zero with some gradients being slightly negative. In addi- The above results show the strong relation between the solidi-
tion, we observed that the velocity of the liquidus front was faster fication process parameters, the resulting structure and the
than that of the solidus front, which increased the size of the mushy mechanical and corrosion properties of the directionally solidified
zone. alloys.
We have also previously found that the recalescence or thermal The aims of the present research were to investigate the role of
arrest and negative gradients in the transition zone are observed. the type of grain macrostructure (columnar, CET and equiaxed) and
After the transition, the speed of the liquidus front is much faster microstructure (secondary dendrite arm spacing) in the wear and
than that of the solidus front. The size of the equiaxed crystal is electrochemical behavior of hypereutectic ZneAl alloys (Zn-1wt.%
always smaller in the transition zone and then normally increases. Al, Zn-2wt.%Al, Zn-3wt.%Al and Zn-4wt.%Al) directionally solidified
The average secondary spacing increases from columnar to equi- in a vertical upward directional solidification device and to corre-
axed zones in the samples. There is a linear and inverse correlation late the thermal parameters with the electrochemical and wear
between the yield strength (YS), maximum tensile strength (TS) properties.
and Vickers microhardness (VH) and secondary dendrite arm
spacing (l2), independently of the type of grains (columnar or
equiaxed). YS, TS and VH depend differently on the grain size
depending on the region considered: columnar or equiaxed.
Rosenberger et al. determined the wear resistance of Zne1wt.%
Al and Zn-2wt.% alloys using a pin-on-ring wear tests, as a function
of type of structure of the as-cast alloys. The three regions were
observed; columnar, equiaxed and CET. The results also showed
that the equiaxed structure is the most resistant. In addition, the
relation between wear rate and grain size depends on the type of
structure and that it is more pronounced for the equiaxed structure.
The effect was almost negligible for the columnar grain structure
[62].
In previous works, we correlated the effect of several parame-
ters, like thermal and metallurgical ones, with electrochemical
parameters on the CET macrostructure in Zn-4wt.%Al, Zn-16wt.%Al
and Zn-27wt.%Al alloys [63]. We were able to observed the
susceptibility to corrosion of the alloys with columnar structure by
analyzing the values of charge-transfer resistance (Rct) obtained
using the Electrochemical Impedance Spectroscopy (EIS) technique.
In Zn-4wt%Al and Zn-27wt.%Al, the corrosion susceptibility
depended on the structure of the alloy. The alloy with 16wt.%Al was
the least resistant to corrosion and its susceptibility to corrosion Fig. 1. Schematic illustration of the water-cooled vertical upward solidification setup.
Fig. 2. Macrostructures: (A) Zn-1wt.%Al, (B) Zn-2wt.%Al, (C) Zn-3wt.%Al and (D) Zn-4wt.%Al, and microstructures of different alloy samples: (1) Columnar, (2) CET and (3) Equiaxed.
398 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414
Fig. 3. (a) Directionally solidified sample of a Zn-1wt.%Al alloy showing the CET. (b) Pin
preparation from columnar zone for wear test. (c) Pin position during the test. (d)
Aspect of the pin after the wear test.
2. Experimental
2.1. Alloy preparation and directional solidification process
Experiments were performed with ZneAl alloys prepared with
four different solute contents; 1wt.%Al, 2wt.%Al, 3wt.%Al and 4wt.%
Al. The chemical compositions of the commercially pure metals
used to prepare the alloys are presented in Table 1. The molds were
made from a 23 mm i.d. and 25 mm e.d. PYREX (Corning Glass
Works, Corning, NY, USA) tube, with a flat bottom, a cylindrical
uniform section and a height of 200 mm. The sample was a cylinder
22 mm in diameter and 100 mm in height.
The alloys were melted in a resistance-type furnace. The
experimental setup consisted of a heat unit, a sample moving
system, a heat extraction system, a temperature control and data
acquisition systems (see schematic representation of Fig. 1). In
Fig. 4. (a) Test electrodes with different type of structures (columnar, equiaxed and CET). (b) Scheme of the position of the test electrode during the electrochemical tests and
(c) Cell kit.
order to understand the evolution of the thermal parameters
during the solidification process, it was necessary to install ther-
mocouples. One K-type thermocouple was used to control the
temperature of the resistance furnace during the experiment and
five K-type thermocouples were positioned at 20 mm intervals in
the centerline of the cylinder mold from the bottom surface of the
mold and were connected to a data acquisition system to record the
temperature history of cooling.
The experimental procedure was as reported previously [15].
First, the liquid metal in the mold in the furnace was allowed to
reach the selected temperature above the melting point of the alloy.
Then, the furnace power was turned off and the melt was allowed
to solidify from the bottom. Heat was extracted through a cooling
system, which consisted of a copper disk attached to a copper coil,
both cooled by running water. The solidification velocity was
adjusted by changing the temperature and water flow and also by
adding plates of materials between the copper plate and the
crucible which changed the effective value of the thermal
conductivity. The crucible was also isolated on the top to reduce
heat losses from the top of the furnace to a minimum. With this
experimental setup, unidirectional heat flow was achieved and the
convection associated with the pouring of the liquid into the mold
was eliminated.
Due to the formation of high pore density at the top of the cast
sample and the formation of a shrinkage pipe, which increase with
the aluminum concentration as reported by Barnhurst [65,66], the
diluted concentrations of ZneAl alloys selected for this study were:
Zn-1wt.%Al, Zn-2wt.%Al, Zn-3wt.%Al and Zn-4wt.%Al. Also, for the
same reason, only the first 100 mm height of the cast was consid-
ered for the sampling. After solidification, the cylinder samples
were cut in the axial direction. One half was used to reveal the
macrostructure and analyze the microstructure, whereas the other
half was cut and machined in coupons 20 mm in length for the
wear and corrosion tests.
2.2. Metallography
In order to reveal the macrostructure, the ZneAl alloys were
polished and etched using concentrated hydrochloric acid for
approximately 30 s at room temperature, followed by rinsing and
 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414
Fig. 5. Temperature versus time curves during directional solidification at different
positions with respect to the chill.
wiping off the resulting black deposit. The selected surfaces were
etched with a mix containing chromic acid (50 g Cr2O3; 4 g Na2SO4
in 100 ml of water) for 15 s at room temperature to reveal the
microstructures [67]. As examples, the representative macro- and
microstructures obtained in the different ZneAl alloys are shown in
Fig. 2. The numbers (1), (2) and (3) in each micrograph correspond
to columnar, CET and equiaxed zones respectively.
The grain size (Gs) and secondary dendritic arm spacing (l2)
were measured at each location using an ArcanoÒ metallographic
microscope and the Soft Imaging Solutions (former SIS) for digital
image analysis and digital image management. The software offers
the functions necessary for acquiring, further processing, interac-
tive/automatic analysis, archiving and documentation of the digital
Fig. 6. Scheme (a) showing the face and interface in a dendrite tip, (b) showing the liquidus and solidus interphases in the sample and (c) showing a sample with the positions of
thermocouple and the CET.
399
images and related investigation results. The secondary dendrite
arm spacings were measured by counting the number of branches
along a line of known length [68]. The mean value of l2 was
calculated from 15 measurements in each zone of the samples
(columnar, equiaxed and CET).
To measure the equiaxed grain size from the CET zone to the top,
each sample was divided into equal intervals. In each interval of
approximately 10 mm, the average diameter of equiaxed grains was
calculated according to the ASTM 112-96 standard norm [69]. The
columnar region was divided into similar size intervals and the
width and length of the columnar grains were measured directly
from the samples.
2.3. Microhardness
Vickers microhardness was measured at room temperature
using loads of 500 gf. The measurements were performed under
standard ASTM E 384-89 [69], using a pressing time of 15 s. The
microhardness of each sample was determined along the longitu-
dinal axis and every 20 mm by taken an average of 10 readings in
each section of the sample.
2.4. Wear tests
The wear tests were performed in a pin-on-ring machine, con-
sisting of a gray cast iron disc of 280 HV in hardness and 170 mm in
diameter, which rotates giving a tangential velocity of 2.7 m s1.
According to the pin-on-ring configuration, the disc axis is hori-
zontally orientated and the pin is mounted vertically on top of the
disc, on the curved edge. Fig. 3(aed) shows the details of the wear
tests.
Cylindrical pins 6.35 mm in diameter and 16 mm in length were
used as samples. The pins were obtained from the casting ingots by
400 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414

performed at room temperature and the ambient moisture ranged

a 3 from 60 to 75% relative humidity.
LIQUID COOLING RATE (K/s)

The temperature of the samples was measured with a thermo-

2.5 couple inserted perpendicularly to the pin axis at 6 mm of the
contact surface. In all the samples, the temperature at 6 mm of the
2 surface was of 70  C on average. The aspect of the pin after the wear
TCET = 1.62 K/s test is presented in Fig. 3(d).
1.5 average position

2.5. Corrosion tests

1
For corrosion testing, samples of 20 mm in length of each zone
0.5 (columnar, equiaxed and CET) and for each concentration were
Zn-1wt%Al prepared as working electrodes using the face which is parallel to
0 the growth direction as active surface. This surface was polished
0 20 40 60 80 100 with sandpaper (from CSi #200 to #1200) and washed with
POSITION (mm) distilled water and dried by natural flow of air. The test electrodes
b of the different structures are shown in Fig. 4(a)
10
All the electrochemical tests were conducted in 3%NaCl solution
VL Zn-1wt%Al
9
(pH ¼ 5.5) at room temperature (25  C) using an IM6d ZahnerÒ
8 VS electrik potentiostat coupled to a frequency analyzer system.
VELOCITY (mm/s)

7 A conventional three-compartment glass electrochemical cell,

with its compartments separated by ceramic diaphragms, was
6
used. The potential of the working electrode (area y 20 mm2) was
5
measured against a saturated calomel reference electrode (0.242 V
4 vs SHE), provided with a Luggin capillary tip. A large area Pt sheet
3 was used as a counterelectrode. A scheme of the electrochemical
VLC = 2.05 mm/s device showing the position of the test electrode during the elec-
2
trochemical tests is presented in Fig. 4(b)
1
Cyclic Voltammetry and Electrochemical Impedance Spectros-
0 copy were used as the principal electrochemical techniques. Vol-
c 0 20 40 60 80 100 tammograms were run at 25  C between preset cathodic (open
POSITION (mm) circuit potential z 1.500 V) and anodic (Es,a ¼ 0.700 V)
switching potentials a potential sweep rate (v) of 0.002 V s1.
T T T T T
1.2
TEMPERATURE GRADIENTS (K/mm)

G 1-2 (Columnar Zone)

Impedance spectra were recorded in the frequency range of
103Hz  f  105 Hz, at open circuit potential.
1 G2
2-3
3 (Columnar
(Co
C lumnar Zone)

G 3-4 (CET Zone)

0.8 For comparison purposes, experiments using pure metals and
G 4-5 (Equiaxed Zone)
aluminum-based alloys with different structures were conducted
G 5-6 (Equiaxed Zone)
0.6 under the same experimental conditions.
0.4
3. Results and discussion
0.2
3.1. Columnar-to-equiaxed transition
0
Gc = 0.11 K/mm
-0.2
Typical macro- and microstructures showing the columnar-to-
0 400 800 1200 1600 2000 2400 2800 3200 3600 4000 4400 4800 5200 5600 equiaxed transition (CET) of the four alloy samples studied are
TIME (Seconds) presented in Fig. 2.
The position of the transition was located by visual observation
Fig. 7. (a) The liquid cooling rate versus distance from the end chill, (b) variation in VL
and optical microscopy and the distance from the base of the
and VS and (c) temperature gradients as a function of time during the end chill
solidification test for one Zn-1wt%Al alloy. sample was measured with a ruler. Fig. 2 shows that the CET is not
sharp, showing a zone delimited by CETMax and CETMin, where
some equiaxed grains co-exist with columnar grains. Normally, the
size of the transition zone is in the order of up to 10 mm.
selecting one sample of each region (equiaxed, columnar or cet) Vanyoussefi and Greer [71] obtained a CET zone in Al-4.15wt%Mg
(see Fig. 3(a) and (b)). The pins were cut with a handsaw and then alloys, whereas Gandin [72] and Ziv and Weinberg [8] found that
lathed at a slow velocity and low stress. One of the plane surfaces of the CET in their experiments was sharp.
the pin, which was ground and polished using SiC abrasive paper
from #220 up to #1500 in grid size and then finished with 0.25 mm 3.2. Determination of solidification parameters
alumina suspensions, was used as the contact surface.
The disc axis was horizontally oriented and the pin was The temperature at five positions of each sample was measured
mounted vertically on top of the disc. A load of 32.4 N was applied by the inserted thermocouples and, as an example, a typical time-
on the pin axis [70]. The total distance and amount of wear were dependent temperature plot for the five thermocouples is shown in
recorded during the test. The wear was measured by the change in Fig. 5 for the Zn-1wt.%Al alloy. The distance from the chill and
length of the pin every 100e200 m of test up to a final distance of structure of each thermocouple position is indicated in the figure.
2000 m. The final distance made sure that the steady state of wear In all five curves it is possible to identify a period corresponding
was achieved in all the tests [66]. All the experiments were to the cooling of the melt, a second period of solidification, and
 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414 401

Table 2
Alloy system and observations of the calculations of thermal parameters at the CET during the directional solidification process.

Test N Alloys (wt%) Minimum CET Maximum CET Cooling rate Critical Critical Averages values of parameters
and liquidus position, position, CETMax. in the melt, liquidus temperature for each composition
and solidus CETMin. (mm) (mm) T_ CET, at the velocity, gradient,
temperatures CET (K s1) VLC (mm s1) Gc (K mm1)
1 Zn-1wt.%Al (ZA1) 31.5  0.1 47.1  0.1 1.6  0.2 2.1  0.1 0.00  0.05
2 58.6  0.1 77.4  0.1 2.3  0.2 1.9  0.1 0.11  0.05 CETAverage ¼ 53.14 mm  0.1
3 TLiquidus ¼ 689.8 K 49.4  0.1 62.8  0.1 2.2  0.2 1.7  0.1 0.26  0.05 T_ CET (Average) ¼ 2.09 K s1  0.2
4 TSolidus ¼ 655 K 43.5  0.1 50.5  0.1 1.9  0.2 1.9  0.1 0.05  0.05 VL(Critical) ¼ 1.91 mm s1  0.1
5 51.0  0.1 59.6  0.1 2.2  0.2 1.7  0.1 0.31  0.05 GL(Critical) ¼ 0.082 K mm1  0.05
6 Zn-2wt.%Al (ZA2) 53.9  0.1 64.5  0.1 2.3  0.2 2.2  0.1 0.15  0.05
7 51.8  0.1 76.1  0.1 2.2  0.2 1.9  0.1 0.19  0.05 CETAverage ¼ 52.57 mm  0.1
8 TLiquidus ¼ 685.2 K 34.7  0.1 48.9  0.1 1.5  0.2 1.5  0.1 0.23  0.05 T_ CET (Average) ¼ 1.93 K s1  0.2
9 TSolidus ¼ 655 K 42.3  0.1 61.6  0.1 1.8  0.2 2.4  0.1 0.38  0.05 VL(Critical) ¼ 2.04 mm s1  0.1
10 39.6  0.1 52.3  0.1 1.6  0.2 1.9  0.1 0.54  0.05 GL(Critical) ¼ 0.222 K mm1  0.05
11 Zn-3wt.%Al (ZA3) 34.7  0.1 39.9  0.1 1.7  0.2 1.8  0.1 1.16  0.05
12 39.4  0.1 43.6  0.1 1.8  0.2 2.0  0.1 0.83  0.05 CETAverage ¼ 33.31 mm  0.1
13 TLiquidus ¼ 671.2 K 31.8  0.1 52.3  0.1 1.5  0.2 2.2  0.1 0.76  0.05 T_ CET (Average) ¼ 1.59 K s1  0.2
14 TSolidus ¼ 655 K 20.9  0.1 34.3  0.1 1.6  0.2 2.0  0.1 0.61  0.05 VL(Critical) ¼ 2.09 mm s1  0.1
15 12.7  0.1 23.5  0.1 1.1  0.2 2.2  0.1 0.59  0.05 GL(Critical) ¼ 0.546 K mm1  0.05
16 Zn-4wt.%Al (ZA4) 23.9  0.1 38.1  0.1 1.6  0.2 2.0  0.1 1.12  0.05
17 28.4  0.1 39.2  0.1 1.5  0.2 2.1  0.1 0.91  0.05 CETAverage ¼ 27.60 mm  0.1
18 TLiquidus ¼ 663.5 K 17.3  0.1 29.8  0.1 1.3  0.2 2.2  0.1 2.01  0.05 T_ CET (Average) ¼ 1.44 K s1  0.2
19 TSolidus ¼ 655 K 24.5  0.1 33.6  0.1 1.4  0.2 2.3  0.1 1.29  0.05 VL(Critical) ¼ 2.16 mm s1  0.1
20 14.9  0.1 26.3  0.1 1.2  0.2 2.0  0.1 0.36  0.05 GL(Critical) ¼ 1.138 K mm1  0.05

a final period of cooling of the solid to ambient temperature. Fig. 5
also shows the cooling rate of the melt, the temperature gradient in
the liquid and the velocities of the isotherm fronts for both the
liquidus and solidus temperatures.
The cooling rate (T) _ was determined from the temperature
versus time curves at each thermocouple position and taking the
average slope of the temperature versus time curve during the
cooling of the melt, before a period of solidification. The tempera-
ture interval used to calculate the average slope was between the
liquidus temperature and the highest temperature above the liq-
uidus reached by the melt before the furnace was turned off.
In the particular case of the interphase velocities, it is necessary
to point out that in the present research we tracked the moving
interphases using thermocouple measurements, and did not track
dendrite tips, since temperature measurements were determined
from a small volume and not from a surface. Temperature
measurements were used to track averaged [L/(S þ L)] interphases
(or liquidus interphases) and [(S þ L)/S] interphases (or solidus
interphases) but not solid/liquid “interfaces”, so dendrite tip
surfaces, or equiaxed grain surfaces may be in any solid þ liquid
region [59]. We assumed that solid can not exist beyond the [L/(S þ
Fig. 8. ZneAl phase diagram [69].
L)] interphase and that liquid can not exist beyond the [(S þ L)/S]
interphase. It is only into the mushy (solid þ liquid) zone where all
the “interfaces” do exist, and the mushy zone is at temperatures
between the local liquidus temperature and the local solidus (or
eutectic) temperature (see Fig. 6).
Each point in Fig. 7(a) corresponds to the liquid cooling rate
value at each thermocouple position in the sample; the white
points correspond to the values in the columnar zone, the black
point to the CET zone and the gray points to the equiaxed zone.
The position of the solidification fronts versus time is deter-
mined by the start and the end of solidification at each thermo-
couple, which correspond to the liquidus and the solidus
temperatures, respectively. Both points are detected by the changes
in the slopes of the cooling curve at the start and end of solidifi-
cation. The velocities of the solidification fronts were calculated as
the distance between the thermocouples divided by the time taken
by either the liquidus or solidus temperature to go from the lower
to the upper thermocouple.
These velocities are denoted as VL and VS, respectively. The
results are shown in Fig. 7(b). The origin of time in the figure
corresponds to the initiation of solidification at the thermocouple
of the bottom of the sample. We can see that the velocity of the
liquid interphase (we did not measure the dendrite tip velocity)
was zero at the beginning of solidification because all is in the liquid
state. It can be noted that after 20 mm the liquid interphase
advances very quickly. On the other hand, the solid front moves
well behind the liquid front and accelerates much less than the
liquid front.
The temperature gradient at all times was calculated by dividing
the temperature difference between two thermocouples by the
separation distance between them. The temperature gradients
calculated from the readings of five thermocouples for Zn-1wt%Al
are presented in Fig. 7(c), and referred to as G12, G23, G34, G45
and G5-6. The dashed vertical lines in the figure show the time at
which the liquid isotherm reaches each thermocouple position;
line T1 coincides with the vertical axis and corresponds to the origin
of the time axis.
It can be observed that the liquidus isotherm reaches the T2
thermocouple position in 1980 s and the T3 thermocouple position
Fig. 9. SEM (Scanning Electron Microscopy) and EDXA (Energy Dispersive X-Ray Microanalysis) of ZneAl alloys showing the distribution of elements. The Zn is distributed in the
b dendrites and in the eutectic phase and a (Al) is present in the eutectic phase (a þ b). Also, it is possible to see the typical morphology of hypoeutectic ZneAl alloys: (a) Zn-1wt.%Al
alloy showing the b (Zn) dendrites without the formation of interdendritic zone. Some porosity is present in these samples, as seen in the photograph. (b) Zn-2wt.%Al alloy and (c)
Zn-3wt.%Al alloy showing the b (Zn) dendrites and a small interdendritic zone with eutectic formation. (d) Zn-4wt.%Al alloy showing the eutectic lamellar morphology (b þ a).
A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414 403

Fig. 9. (continued).
404 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414

Fig. 9. (continued).
A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414 405

Fig. 9. (continued).
406 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414
in 2260 s, which is the time when the transition at T3 occurs. At the
beginning, the gradients have values around 1 K mm1 at the
bottom of the sample, and 0.4 K mm1 at the top. These values
decrease gradually, when the transition occurs, and the gradient in
the liquid becomes small, nearly to zero or slightly negative (about
0.1 K mm1 in this specific case). This is the value of the critical
temperature gradient, GL(Critical), for this experiment. This zero or
slightly negative value of thermal gradient in the liquid at CET was
first proposed by Ch.-A. Gandin [14] and also reported by us [13]. A
similar behavior can be observed in all the samples. The critical
temperature gradients for all the experiments are listed in Table 1.
The scatter in the values of gradient is associated with the fact
that the transition usually occurs between two thermocouples and
therefore, the calculated value is an average over a region which
includes the mushy zone and the melt. In addition, according to the
relative error presented above, at these low values of temperature
differences, the error in the calculated gradients at these low values
of temperature differences could be as large as 45%.
Thereby, taking these considerations into account, it can be
concluded that within the error, the alloy composition does not
have an effect on the temperature gradient during the transition.
Comparing the separation in time (x axis) between the dashed
vertical lines in Fig. 7(c), which determines the time when the
liquidus isotherm reaches the position of the corresponding ther-
mocouple, it is observed that there are 1980 s between T1 and T2,
280 s between T2 and T3, 240 s between T3 and T4 and 120 s between
T4 and T5. This shows that there is an acceleration of the liquidus
interface and that the transition occurs not only at a low gradient
but also with large interface velocities.
The most important thermal parameters extracted from the
temperature versus time curves at the CET during the directional
solidification process are listed in Table 2 for each hypoeutectic
alloy.
3.3. Microstructures formed in hypoeutectic alloys during
directional solidification
According to the ZneAl phase diagram (Fig. 8) the phase formed
at 418  C is b (Zn) and liquid (L) up to 382  C. Below this temper-
ature, part of b (Zn) forms the g phase (ZnAl) þ b (Zn), up to 275  C,
when the eutectoid transformation g (ZnAl) / a (Al) þ b (Zn)
occurs. Below 275  C, b (primary, rich in Zn) þ eutectoid (a þ b) is
formed. However, our directional solidification is a non-equilibrium
solidification, and then high-temperature phases can be retained
towards room temperature. Fig. 9(a) shows the microstructure of
the Zn-1wt.%Al alloy, in the equiaxed zone, showing the b (Zn)
dendrites without the formation of interdendritic zone.
In the case of the hypoeutectic alloys Zn-2wt.%Al, Zn-3wt.%Al
and Zn-4wt.%Al, at approximately 416  C, the b (Zn) phase þ liquid
(L) is formed up to 382  C, when the eutectic transformation L /
b (Zn) þ g (ZnAl) occurs. Below 382  C, b (Zn) phase þ Eutectic (b
(Zn) þ g (ZnAl)) is formed up to 275  C, when the eutectic trans-
formation g (ZnAl) / a (Al) þ b (Zn) occurs. Below 275  C, the
b (primary, rich in Zn) þ eutectoid (a þ b) is formed.
From Zn-2%wt.Al to Zn-4%wt.Al alloys, the eutectic phase (inter-
dendritic region) formed increases at expense of the b (Zn), with
k ¼ CS/CL ¼ 0.20 (see Table 3 and Fig. 9bed
3.4. Evolution of structural parameters with directional
solidification
The typical EDXA micrograph of the hypoeutectic and hyper-
eutectic ZneAl alloys studied is presented in Fig. 9. From Zn-2wt%Al
(Fig. 9(b)) to Zn-4wt.%Al (Fig. 9(c)) it is possible to observe that the
dendritic matrix is Zn-rich (b phase) whereas the interdendritic
region, which is the eutectic (a þ b) phase, is Al-rich.
We carefully analyzed the microstructures of all the samples
using EDXA and found no evidence of the presence of the discon-
tinuous precipitation (DC) phenomenon in the samples, at room
temperature and along the time during which the experiments
were performed [49e58]. The presence of porosity was observed in
some samples, as seen in Fig. 9(a).
Fig. 10 shows the evolution of the secondary dendrite arm
spacing (l2) as a function of the distance from the caloric extraction
surface (bottom of the sample) for all the hypoeutectic alloys
studied (Zn-1wt%Al, Zn-2wt%Al, Zn-3wt%Al and Zn-4wt%Al). It is
possible to see that l2 increases from the columnar to equiaxed
zone of the samples, because the heat extraction and cooling rate
are higher during columnar than during equiaxed growth,
producing a shorter solidification time.
It is also observed that l2 decreases from 1wt%Al to 4wt%Al
alloys, which is expected for the higher increase in the solute at
dendrite interface arms, which is predicted by models [73]. In
addition, there might be a difference in the solidification time
between both experiments. The values of l2 versus position (P,
distance from the bottom) are fitted byl2 ¼ a:P b, where the value
of parameter a decreases from 38.33 (Zn-1wt%Al) to 14.43 (Zn-4wt
%Al) and the value of b increases from 0.16 (Zn-1wt%Al) to 0.26 (Zn-
4wt%Al), see Table 5.
The evolution of grain size with distance from the bottom of the
sample is presented in Fig. 11.
The size of the equiaxed grains for the Zn-1wt%Al sample is
1.5 mm in the transition region and then starts to monotonically
increase up to a value of 4.5 mm at the end of the sample. In the
case of the width of the columnar grains, it is observed that the
size increases from 2 mm in whole columnar zone to 3.8 mm at
the end of the transition region. A similar analysis was performed
for all solidification experiments, as shown in Fig. 11 for other
alloy concentrations (the equiaxed points in the figures were
fitted to polynomial functions of third degree). It is observed that
in the transition region, the size of the equiaxed grains is smaller
than the width of the columnar grains (the black points in the
figures correspond to the measurements of the size of the
columnar and equiaxed dendrites in the CET zone). However, in
all cases, the size increases after the transition. At the end of
solidification, the size of equiaxed grains also reaches a maximum
value.
The figures show that the grain size in the Zn-2wt%Al sample
increases from 1.8 mm to 2.7 mm in the columnar zone and from
1.4 mm to 3.8 mm in the equiaxed zone. In the Zn-3wt%Al alloy, the
grain size is between 1.7 mm and 2.7 mm in the columnar area and
between 1.2 mm and 3.5 mm in the equiaxed area. Finally, for the
Zn-4wt%Al alloy, the size increases from 1.3 mm to 3.2 mm
(columnar zone) and from 0.8 mm to 2.2 mm (equiaxed zone). The
size of the grains in the three regions (columnar, equiaxed and CET)
decreases from the Zn-1wt%Al to the Zn-4wt%Al alloy.
The values of grain size (GS) versus position (P) are fitted by
GS ¼ c:P d . For the columnar zones, the value of c varies from 1.17
(Zn-1wt%Al) to 0.79 (Zn-4wt%Al) and the value of d increases from
0.25 (Zn-2wt%Al) to 0.31 (Zn-4wt%Al), whereas for the equiaxed
zones the value of c increases from 0.01 (Zn-1wt%Al) to 0.12 (Zn-
4wt%Al) and the value of d decreases from 1.27 (Zn-2wt%Al) to 0.58
(Zn-4wt%Al), see Table 6.
The evolution of l2 with GS is shown in Fig. 12 for all alloy
concentrations. In both cases, when the grain size increases, l2
increases as well in the columnar and equiaxed zones of the
samples. The values of both l2 and grain size of columnar grains in
the area of the CET are usually larger or approximately equal to the
size in the equiaxed zone.
 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414 407

Table 3 100
Solid and liquid fractions of hypoeutectic ZneAl alloys. 90 • Equiaxed
Hypoeutectic ZneAl alloy fs (Dendritic region) fl (Interdendritic region) 80 • CET
70 о Columnar
Zn-1% wt.Al 87% 13%

λ 2 (μ m)
Zn-2%wt.Al 68% 32%
60
Zn-3% wt.Al 47% 53% 50
Zn-4%wt.Al 25% 75% 40
30
20
10
It is observed that there is a relation that indicates that the grain 0
size increases as l2 increases in both the columnar and equiaxed 0 10 20 30 40 50 60 70 80 90 100
regions. However, as shown in Fig. 12, the relation is different in POSITION (mm)
each zone, showing that a given grain size in the columnar zone is
formed with dendrites of smaller l2 than in the equiaxed zone. It is Zn-1w t%Al Zn-2w t.%Al Zn-3w t.%Al Zn-4w t.%Al
necessary to mention that what is reported as grain size in the
columnar zone is actually the columnar width. Also, the local
Fig. 10. Grain size experimentally measured as a function of distance from the bottom
solidification time is higher in the CET and equiaxed zones than in of the samples.
the columnar zone (see Fig. 5), contributing to the increase in grain
size and dendrite spacing in the equiaxed zone.
The values of secondary l2 versus grain size (GS) are fitted by As the concentration of aluminum in the alloy increases, the
l2 ¼ e:GfS . For the equiaxed zones, the value of e decreases from definition of these peaks of reduction is not clear, although the C2
69.45 (Zn-1wt.%Al) to 36.95 (Zn-4wt.%Al) and the value of f peak is dominant. The most important difference is observed in the
increases from 0.10 (Zn-1wt.%Al) to 0.45 (Zn-4wt.%Al), whereas for distribution of the cathodic current peaks, which indicates the
the columnar zones, the value of e decreases from 35.47 (Zn-1wt.% different characteristics of the films formed during the anodic scan.
Al) to 17.78 (Zn-4wt.%Al) and the value of f increases from 0.49 (Zn- These results can be attributed to the aggressive/depassivating
21wt.%Al) to 0.91 (Zn-4wt%Al), see Table 7. action of Cl anions [74]. At present, the mechanism of film
formation is still uncertain.
For the case of CET, profiles are more complex, because the
3.5. Corrosion resistance of different alloys and structures
proportion of one or other type of grains (columnar or equiaxed)
can vary from sample to sample (Fig. 13(c)).
3.5.1. Voltammetric data
For the equiaxed structures and for more diluted alloys (Zn-
Fig. 13 shows the voltammograms of the columnar and equiaxed
1wt.%Al) the reduction current decreases to about half, indicating
zinc (Fig. 13(a)) and of different hypoeutectic alloys and structures,
that the passive layer formed on the material is more protective
that is, columnar, CET and equiaxed (Fig. 13bed) In all the cases, it is
(Fig. 13(d)).
possible to observe that during the anodic potential scanning the
In the alloys, the values of the currents are similar for the same
current is practically zero until it reaches the potential of 1 V,
Al concentration and the most important difference is observed in
when the current increases sharply, thus beginning the active
the cathodic current peaks, which indicates the different charac-
dissolution of metal. The voltammetric curves in the backward
teristics of the films formed during the anodic scan. As it is known,
direction show a hysteresis loop, suggesting that this current
the aluminum corrosion products are more insulated than those of
increase was due to the start of a process of pitting. Two cathodic
zinc.
current peaks at about 1.2 V and 1.3 V (called C1 and C2,
The orientation of the grains in a polycrystalline zinc alloy has
respectively) for the equiaxed structure in Fig. 13(a)) are also
preferential orientations depending on the casting process and the
present, which could be associated with the reduction of ZnO and
mechanical working conditions. For example, in the case of cast
Zn(OH)2 [37]. In the case of columnar structure only one C1 peak
products the <0001> direction is perpendicular to the axis of the
corresponding to ZnO is observed.
cast columnar crystals [75]. Thus, in the columnar sections, the
grains grow in more favorable directions to promote the formation
Table 4
Values of the Electrochemical Impedance Spectroscopy (EIS) adjusting parameters
of a film with more protective characteristics.
for each structure and alloy concentration. This different contribution of the peaks in the voltammograms
gives rise to surface layers with different corrosion products, where
Alloy and type of structure Rct (U cm2) C (F cm2)
samples with 4wt.%Al show the formation of a thicker layer of
Zn Columnar 414 6.4  1006  1*107
corrosion products (see Fig. 14aed for Zn-1wt.%Al, Zn-2wt.%Al, Zn-
Zn CET 401 3.8  1006  1*107
Zn Equiaxed 384 2.8  1006  1*107 3wt.%Al and Zn-4wt.%Al respectively).
Al Columnar 19 3.3  1006  1*107
Al CET 15 1.7  1005  1*106 3.5.2. Electrochemical impedance spectroscopy data
Al Equiaxed 11 2.7  1004  1*105
Impedance spectra are dependent on the composition and
Zn-1wt.%Al Columnar 983  10 1.9  105  1*106
Zn-1wt.%Al 358  10 3.0  104  1*105
structures of the alloys. The experimental Nyquist diagrams of the
CET
Zn-1wt.%Al Equiaxed 310  10 2.4*104  1*105
Zn-2wt.%Al Columnar 800  10 1*104  1*105 Table 5
Zn-2wt.%Al CET 503  10 2.4*104  1*105 l2 versus position (P) fitted byl2 ¼ a:P b : values of a and b for different alloys and
Zn-2wt.%Al Equiaxed 300  10 3.2  104  1*105 structures.
Zn-3wt.%Al Columnar 590  10 1.3*104  1*105
Zn-3wt.%Al 390  10 2.1*104  1*105 Alloy l2 versus position R2
CET
Zn-3wt.%Al Equiaxed 234  10 3.6  104  1*105 Zn-1wt.%Al l2 ¼ 38:33:P 0:1628 0.9448
Zn-4wt.%Al Columnar 351  10 8.6  104  1*105 Zn-2wt.%Al l2 ¼ 31:33:P 0:1718 0.9682
Zn-4wt.%Al CET 345  10 1.1*104  1*104 Zn-3wt.%Al l2 ¼ 25:44:P 0:1971 0.9530
Zn-4wt.%Al Equiaxed 180  10 1.7*103  1*104 Zn-4wt.%Al l2 ¼ 14:43:P 0:2660 0.9773
408 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414

5 100
• Equiaxed •
90
•
4 • CET 80 о
GRAIN SIZE (mm)

о Columnar
70
60

λ (μ m)
3
50
40
2
30
20
1
10
0
0 0 1 2 3 4 5
0 10 20 30 40 50 60 70 80 90 100 GRAIN SIZE (mm)
POSITION (mm)

Zn-1w t%Al Zn-2w t.%Al Zn-3w t.%Al Zn-4w t.%Al

Fig. 12. Secondary dendrite arm spacing (l2) versus grain size in different zones of the
Fig. 11. Secondary dendritic arm spacing versus grain size in different zones of the samples.
samples.

Zn-1wt%Al, Zn-2wt%Al, Zn-3wt%Al and Zn-4wt%Al alloys for the
different structures are presented in Fig. 15aed respectively. The
diagrams show one capacitive time constant at high and interme-
diate frequencies and a not well-defined time constant at low
frequencies.
For the Zn-1wt.%Al, Zn-2wt.%Al and Zn-3wt.%Al alloys, the
impedance values depend on the alloy structure and only for the
equiaxed sample, the capacitive loop at high frequencies is well
defined. On the other hand, as the Al concentration increases to
4wt.%Al, the impedance data are similar for all the structures.
The whole set of experimental impedance spectra can be dis-
cussed according to the following total transfer function, corre-
sponding to the circuit of Fig. 16:
Zt ðjwÞ ¼ RU þ Z
with:
1 1
¼ þ j:w:C
Z Rct þ ZW
where RU is the ohmic solution resistance, u ¼ 2pf; Cdl the
capacitance of the electric double layer, Rct the charge transfer
resistance and ZW the diffusion contributions in impedance spectra.
The good agreement between the experimental and simulated
data according to the transfer function given in the analysis of Eqs.
(1) and (2) using non-linear least square fit routines is demon-
strated in Fig. 17 for the Zn-1wt.%Al alloy (CET structure). The values
of C and Rct determined from the optimum fit procedure are pre-
sented in Table 4.
The analysis of the impedance parameters associated with the
time constant at low frequencies is difficult because it is incomplete
in some cases. However, it was possible to calculate an approximate
value of diffusion coefficient D z 1010e1012 cm2 s1 from the
fitting.
The corrosion current can be related to the Rct in the case of
mixed control [76], where the polarization resistance technique
fails, according to the following expression:
Rct ¼ ba :bc =½2:303:ðba þ bc Þ:Icorr  (3)
The experimental data shown in Table 4 indicate that the values
of Rct decrease from columnar to equiaxed structures and from
1wt.%Al to 4wt.%Al alloys. These results indicate that when the
aluminum concentration increases in the interdendritic region
(eutectic lamellar mixture) the susceptibility to corrosion increases
and the effect is more notorious for the equiaxed structure.
Furthermore, the Rct values are related to the rate of charge
(1)
transfer reactions that give rise to the formation of a corrosion layer
on the surface of the samples but say nothing about the protective
characteristics of the film formed. The double-layer capacitance,
Cdl, gives us this information. Table 4 shows that Cdl increases from
(2) Zn-1wt.%Al to Zn-4wt.%Al alloys. The high values of Cdl in the case
of Zn-4wt.%Al alloys with equiaxed structure confirm the formation
of porous corrosion products (Fig. 14(d)), as determined by other
authors by X-ray diffraction [77].
3.6. Microhardness
The Vickers microhardness of four different alloys versus posi-
tion with respect to the chill is plotted in Fig. 18. It can be seen that
Vickers microhardness increases from the columnar to equiaxed
zones of the samples as well as with the increase in aluminum
content in the samples.
The directional solidification process creates gradients of
composition along the length of the material. This, in turn, may

Table 6 Table 7
Grain size (GS) versus position (P) fitted by GS ¼ c:P d : values of c and d for different Secondary spacing (l2) versus grain size (GS) fitted byl2 ¼ e:GfS : values of e and f for
alloys and structures. different alloys and structures.

Alloy Grain size versus position R2 Alloy l2 versus grain size R2

Zn-1wt.%Al Columnar GS ¼ 1:166:P 0:3292 0.9939 Zn-1wt.%AlColumnar l2 ¼ 35:467:G0:4953
S 0.9226
Zn-1wt.%Al Equiaxed GS ¼ 0:012:P 1:2732 0.9285 Zn-1wt.%AlEquiaxed l2 ¼ 69:446:G0:101
S 0.7534
Zn-2wt.%Al Columnar GS ¼ 1:216:P 0:2523 0.9814 Zn-2wt.%AlColumnar l2 ¼ 27:242:G0:6921
S 0.8664
Zn-2wt.%Al Equiaxed GS ¼ 0:020:P 1:1145 0.9024 Zn-2wt.%AlEquiaxed l2 ¼ 61:496:G0:0716
S 0.8613
Zn-3wt.%Al Columnar GS ¼ 1:072:P 0:2617 0.9618 Zn-3wt.%AlColumnar l2 ¼ 25:308:G0:6879
S 0.8541
Zn-3wt.%Al Equiaxed GS ¼ 0:013:P 1:1664 0.8038 Zn-3wt.%AlEquiaxed l2 ¼ 55:591:G0:133
S 0.8339
Zn-4wt.%Al Columnar GS ¼ 0:792:P 0:3074 1 Zn-4wt.%AlColumnar l2 ¼ 17:778:G0:9144
S 0.7895
Zn-4wt.%Al Equiaxed GS ¼ 0:127:P 0:58 0.8851 Zn-4wt.%AlEquiaxed l2 ¼ 36:953:G0:4548
S 0.7571
 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414 409

Fig. 13. Voltammograms of (a) Zn equiaxed and (bed) ZneAl diluted alloys with different structures: (b) columnar (c) CET and (d) equiaxed.

affect the hardness and, consequently, the wear resistance. The
composition of the different alloying elements along the cast
samples was analyzed by Scanning Electron Microscopy (SEM) and
Energy Dispersive X-Ray Microanalysis (EDXA) and some results
are shown in Fig. 19aed.
In the macroscale, the aluminum concentration increases in the
direction of solidification from the bottom to the top of the samples.
We found that the silicon concentration also increases but that the
iron composition decreases in the direction of solidification. The
increase in Si is more pronounced with the increase in aluminum
concentration in the alloy, which could explain the increase in
microhardness in the case of Zn-4wt.%Al in the CET and equiaxed
zones with respect to the other alloys with lower concentrations of
Al (Fig. 19).

Fig. 14. Micrographs showing the formation of corrosion products on the surface of different samples with equiaxed structure, after the cyclic voltammetry test. (a) Zn-1wt.%Al, (b)
Zn-2wt.%Al, (c) Zn-3wt.%Al and (d) Zn-4wt.%Al.
410 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414

a c

b d

Fig. 15. Nyquist of (a) Zn-1wt.%Al, (b) Zn-2wt.%Al, (c) Zn-3wt.%Al and (d) Zn-4wt.%Al alloys with different grain structures.

3.7. Wear resistance

The results of the wear rate as a function of ZneAl alloys with

different structures and composition are shown in Fig. 20. From
these results it is possible to calculate the wear resistance as the
mathematical inverse of the wear rate.
It is observed that the wear rate of the equiaxed structure is
lower than that of the columnar structure. When analyzing the
influence of the aluminum composition on the wear rate inde-
pendently of the structure it is observed that as the aluminum
concentration increases, the wear rate decreases.
The gray lines plotted in Fig. 20 show the influence of aluminum
concentration by type of structure. The dark gray line indicates the

Fig. 17. Bode diagram for Zn-4wt%Al indicating the experimental data (.) and the
Fig. 16. Equivalent circuit corresponding to the transfer functions given by Eq. (2). fitted values (d).
 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414 411

55 aluminum concentrations equal to and lower than 2 wt%, the wear

Zn Load = 500 gf
rates are intermediate to the wear rate of the equiaxed and columnar
Zn-1w t%Al
Zn-2w t%Al
structures, whereas at 3 wt.% the wear rate is larger and at 4 wt.%
50
Zn-3w t%Al lower than that of the equiaxed and columnar structures.
Zn-4w t%Al The behavior observed in the equiaxed structure indicates
45 a gradual increase in the strengthening when the amount of alloying
increases, whereas when a constant value is achieved, it indicates
HV

that the wear mechanism has changed and the wear rate is no longer
40
controlled directly be the alloying. On the other hand, the sudden
decrease in the wear rate of the columnar structure when the
35
composition increases from 2wt.%Al to 3wt.%Al indicates a transi-
• Equiaxed zone tion in the wear mechanism. The wear resistance is controlled by the
• CET zone strength of the materials, which is formed by the mechanical
о Columnar zone
30 resistance of the grains and the border of grains. Comparing both
0 10 20 30 40 50 60 70 80 90 100
structures, an equiaxed structure is preferred in those wear condi-
LENGTH OF THE SAMPLE (mm) tions because no sudden transition in wear mechanisms is observed.
All these samples present a discontinuous thin layer of transfer
Fig. 18. Microhardness values versus length of ZneAl alloy samples using a load of
500 gf. material on the contact surface of the pin composed of iron, zinc
and aluminum. This layer is known to influence the wear behavior
wear rate of the equiaxed structure whereas the light gray line [70,78]. This kind of layer has been found by other authors [44,45]
indicates the wear rate of the columnar structure. and this tribolayer has an important role in the wear process
The wear rate of the equiaxed structure decreases from 1.15 x although it has been found that in systems that present this layer,
103 mm m1 to a constant value of 4 x 104 mm m1 when the the wear rate is controlled mainly by the resistance of the subsur-
aluminum concentration increases from 0 to 2wt.%Al, and it is the face of the materials [70,79].
same value up to 4wt.%Al. On the other hand, for the columnar
structure, the wear rate is constant for an aluminum concentration
lower than 3 wt.%, with a value of approximately 1.5 x 3.8. Relationship between corrosion resistance and thermal
103 mm m1, and at larger concentrations it decreases suddenly to parameters
a constant value of approximately 4 x 104 mm m1.
The CET zone shows an irregular behavior. This is for the mixing of Fig. 21(a) shows the correlation between the charge-transfer
columnar and equiaxed structures that is not constant in the different resistance Rct and the critical temperature gradient. It is observed
alloys (the proportion of each structure with respect to another): i.e. at that in the experiments with ZneAl hypoeutectic alloys, the Rct

Fig. 19. Composition of the alloy versus distance from the chill. (a) Zn-1wt.%Al, (b) Zn-2wt.%Al, (c) Zn-3wt.%Al and (d) Zn-4wt.%Al alloys.
412 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414

1.8E-03
a 1200
1.6E-03 Columnar Zn-1w t.%Al

Charge-Transfer Resistance (ohm.cm )

2
WEAR RATE [mm/m]

Equiaxed
1.4E-03
1000 CET
1.2E-03 Zn-2w t.%Al

1.0E-03 800

8.0E-04
600
6.0E-04

4.0E-04 Zn-3w t.%Al

Zn-4w t.%Al
400
2.0E-04

0.0E+00
200
Zn-1%Al-CTE

Zn-2%Al-CTE

Zn-3%Al-CTE

Zn-4%Al-CTE
Zn-CTE

Zn-1%Al-E

Zn-2%Al-E

Zn-3%Al-E

Zn-4%Al-E
Zn-1%Al-C

Zn-2%Al-C

Zn-3%Al-C

Zn-4%Al-C
Zn-E

Zn-C

0
-2 -1.5 -1 -0.5 0

ZINC-ALUMINUM ALLOY AND STRUCTURE Average Critical Temperature Gradient (ºC/cm)

b 1200
Fig. 20. Wear rate in function of the composition of the alloy (Zn-1wt%Al, Zn-2wt%Al,
Columnar
Zn-3wt%Al and Zn-4wt%Al) and structure (columnar, equiaxed and CET). Zn-1w t.%Al

Charge-Transfer Resistance (ohm.cm )

2
Equiaxed
1000 CET
Zn-2w t.%Al

increases with the critical temperature gradient from 4wt.%Al to 800

1wt.%Al alloys. In contrast, the hypoeutectic alloys with the lowest
critical temperature gradients are the least resistant to corrosion.
Fig. 21(b) shows that Rct increases with the increase in the 600
_ from 4wt.%Al to 1wt.%Al alloys. On
average critical cooling rate (T) Zn-3w t.%Al
Zn-4w t.%Al
the other hand, the hypoeutectic alloys with the lower cooling rates 400
are the less resistant to corrosion.
Fig. 21(c) shows the correlation between the variation in Rct and
200
the average liquidus interphase velocity (VLC). It is observed that Rct
diminishes with the increase in VLC from 1wt.%Al to 4wt.%Al alloys.
In contrast, the hypoeutectic alloys with the highest average 0
1 1.5 2 2.5
interphase velocity are the most resistant to corrosion.
Average Critical Cooling Rate (ºC/s)
c 1200
3.9. Relationship between corrosion resistance and structural Columnar
Charge-Transfer Resistance (ohm.cm )
2

parameters Equiaxed
1000 CET

Fig. 22(a) shows the variation in the charge-transfer resistance

Rct versus the solid fraction (fS), representing the dendritic region 800
formed. It is observed that Rct increases with fS from 4wt.%Al to
1wt.%Al alloys. On the other hand, the hypoeutectic alloys with the 600
lowest solid fraction are the least resistant to corrosion.
Fig. 22(b) and (c) shows Rct plotted as a function of grain size
(GS) and secondary dendritic arm spacing (l2). It is possible to see
400

that Rct decreases from the columnar to equiaxed zones of the

samples, as both GS and l2 increase.
Zn-1w t.%Al
200
Zn-2w t.%Al
The values of grain size in the CET zones are smaller than in the Zn-3w t.%Al

equiaxed zones and Rct decreases from Zn-1wt%Al to Zn-4wt%Al 0

Zn-4w t.%Al

alloys (black points inside the circle in Fig. 22(b)). 1.5 2 2.5
The correlation between Rct and l2 shows the same behavior as Critical Liquidus Interphase Velocity (mm/s)
for GS (Fig. 22(c)). Rct decreases with the increase in l2 and from the
columnar to the equiaxed zones of the samples. In this case, the Fig. 21. Correlation between Rct and: (a) Critical temperature gradient (Gc), (b) Cooling
_ (c) Critical liquidus interphase velocity (VLC).
rate (T),
values of the CET zone are between the values of both columnar and
equiaxed zones.

3.10. Relationship between wear rate and structural parameters at lower grain sizes. In contrast, the wear rate of the CET zone
decreases rapidly as the GS decreases to values lower than those
Fig. 23(a) shows the wear rate versus the grain size for the three corresponding to the equiaxed or columnar grains. The decrease in
types of structures (columnar, equiaxed and CET). It is observed the wear rate is attributed to the strengthening by the border grain
that the wear rate increases as the GS increases. theory [69], which establishes that the stacking of dislocation in the
The wear rates of the columnar grains are higher than those of border of grains causes the strengthening of the material.
the equiaxed grains for similar grain sizes. It is remarkable that the Fig. 23(b) shows the wear rate versus average dendritic spacing
wear rate of the columnar and equiaxed structures becomes similar in each zone of the samples of different alloys. The results show
 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414 413

a 1200 a 2.0E-3
Columnar Zn-1w t.%Al
EQ.
Equiaxed
1000
CET

WEAR RATE [mm/m]

Zn-2w t.%Al
CET

800 COL.
Rct (ohm.cm )
2

600

Zn-3w t.%Al
Zn-4w t.%Al
400

200

0.0E+0
0
0 10 20 30 40 50 60 70 80 90 100 0 0.5 1 1.5 2 2.5 3 3.5 4

Solid Fraction (% ) GRAIN SIZE [mm]

b 1200 b 2.0E-3
x Columnar Zn-1wt.%Al
1000 x Equiaxed
EQ.
Zn-2wt.%Al

WEAR RATE [mm/m]

Zn-3wt.%Al
2
Rct / Ohm. cm

800 CET
Zn-4wt.%Al

600
COL.
400

200

0
0 0.1 0.2 0.3 0.4 0.5

c Gs (cm)
1200
Zn-1wt.%Al x Columnar
1000 Zn-2wt.%Al x Equiaxed 0.0E+0
Zn-3wt.%Al 0 10 20 30 40 50 60 70 80 90 100
Rct (Ohm.cm )
2

800 Zn-4wt.%Al
λ 2 [μ m]
600
Fig. 23. Wear rate versus (a) grain size and (b) secondary dendritic spacing for three
400 types of structure (columnar, CET and equiaxed) and for all the compositions together.

200

0
4. Conclusions
0 10 20 30 40 50 60 70 80 90 100

λ 2 (μ m) Here we determined the main parameters which characterize

the columnar-to-equiaxed transition (CET) phenomenon during the
Fig. 22. Charge-transfer resistance (Rct) versus (a) solid fraction (fS), (b) grain size (GS)
and (c) secondary dendrite arm spacing (l2).
directional solidification process as well as the principal parame-
ters which characterize the corrosion and wear resistance of
hypoeutectic ZneAl alloy samples. We also determined the relation
that the wear rate increases as l2 increases, in a stronger relation between thermal, corrosion and wear parameters and can draw the
than in the case of the dependence of wear rate on GS. In addition, following conclusions:
there are clear differences among the types of dendrite (columnar,
equiaxed and CET), indicating that for a given spacing, the wear rate 1. Samples of ZneAl hypoeutectic alloys with CET were obtained
decreases in the same order. This stronger relation between l2 and and the transition occurs in a zone rather than in a sharp plane,
the type of dendrite strongly indicates a more fundamental relation where both columnar and equiaxed grains coexist in the melt.
between wear resistance and dendritic spacing than between wear The thermal parameters during the transition vary as follows:
resistance and grain size. the critical temperature gradient in the melt varies between
The wear rates of the columnar grains are higher than those of 0.08 K mm1 and 1.13 K mm1, the critical liquid isotherm
the equiaxed grains at a same l2, and both converge at similar velocity varies between 1.91 mm s1 and 2.16 mm s1 and the
values when l2 decreases. The wear rate of the CET zone is inter- cooling rate in the melt varies between 1.44 K s1 and 2.09 K s1,
mediate between the columnar and equiaxed grains. with average transition zones between 27.6 mm and 53.1 mm.
In diluted ZneAl alloys it is observed that there is a larger 2. The average secondary spacing increases from columnar to
influence of the structure of the subsurface on the wear resistance, equiaxed zones in the samples and the results of the
in particular when it is related to l2 for each structure. The wear measurements of the grain size show that in the CET region, the
resistance of the CET zone has an intermediate value because it is equiaxed grains are smaller, and that after the transition, in the
a mix of columnar and equiaxed grains and a non-proportional mix equiaxed zone, the grain size either increases or remains
could explain the relation between the wear rate and l2. approximately constant.
414 A.E. Ares et al. / Materials Chemistry and Physics 136 (2012) 394e414
3. From the analysis of the values of charge-transfer resistance,
Zn-1wt.%Al alloys with columnar and equiaxed structures are
the most resistant to corrosion. In the case of the samples with
CET structures, the most resistant to corrosion is Zn-2wt.%Al.
However, the charge-resistant transfer criterion for the evalu-
ation of corrosion resistance of the alloys does not indicate the
protective characteristics of the films formed on hypoeutectic
ZneAl alloys directionally solidified.
4. From the analysis of the current values in the polarization
curves obtained, it is possible to appreciate that, in the case of
hypoeutectic ZneAl alloys, the equiaxed structure is the most
susceptible to corrosion. In the alloys analyzed, the values of
the currents are similar for the same Al concentration and the
most important difference is observed in the distribution of the
cathodic current peaks, which indicates the different charac-
teristics of the films formed during the anodic scan.
5. In the case of the correlation between the charge-transfer
resistance and the structure, the values of Rct decrease from
the columnar to the equiaxed zones of the samples, as both
structural parameters (grain size (Gs) and secondary dendrite
arm spacing (l2)) increase.
6. Rct increases with the increase in GLC, in T_ and in fS, from 4wt.%
Al to 1wt.%Al alloys and decreases with the increase in VLC from
1wt.%Al to 4wt.%Al alloys.
7. For the same wear conditions, the wear rate of the equiaxed
region is lower than that of the columnar and transition
regions. Independently of the type of structure (columnar,
equiaxed or CET), the wear resistance increases with the
aluminum concentration. For each alloy concentration, the
wear resistance increases from the columnar to the equiaxed
structure.
8. A clear inverse relation is observed between the wear resis-
tance and the secondary dendritic spacing, which may explain
the increase in wear resistance in the equiaxed region.
Acknowledgments
We would like to thank the National Science Research Council of
Argentina (CONICET) for the financial support.
References
[1] J.A. Spittle, Int. Mater. Rev. 51 (2006) 247e269.
[2] S.C. Flood, J.D. Hunt, Columnar to Equiaxed Transition, ASM Handbook, ASM
International, Materials Park, OH, 1998.
[3] G. Reinhart, N. Mangelinck-Noël, H. Nguyen-Thi, T. Schenk, J. Gastaldi, B. Billia,
P. Pino, J. Härtwig, J. Baruchel, Mater. Sci. Eng. A 413e414 (2005) 384e388.
[4] S. McFadden, D.J. Browne, C.A. Gandin, Metall. Mater. Trans. A 40 (2009)
662e672.
[5] C.A. Siqueira, N. Cheung, A. Garcia, Metall. Mater. Trans. A 33 (2002)
2107e2118.
[6] B. Willers, S. Eckert, U. Michel, I. Haase, G. Zouhar, Mater. Sci. Eng. A 402
(2005) 55e65.
[7] R.B. Mahapatra, F. Weinberg, Metall. Trans. B 18 (1987) 425e432.
[8] I. Ziv, F. Weinberg, Metall. Trans. B 20 (1989) 731e734.
[9] J.D. Hunt, Mater. Sci. Eng. 65 (1984) 75e78.
[10] S.C. Flood, J.D. Hunt, J. Cryst. Growth 82 (1987) 543e551.
[11] S.C. Flood, J.D. Hunt, J. Cryst. Growth 82 (1987) 552e560.
[12] C.Y. Wang, C. Beckermann, Metall. Mater. Trans. A 25 (1994) 1081e1093.
[13] A.E. Ares, C.E. Schvezov, Metall. Mater. Trans. A 31 (2000) 1611e1625.
[14] C.A. Gandin, Acta Mater. 48 (2000) 2483e2501.
[15] A.E. Ares, C.E. Schvezov, Metall. Mater. Trans. A 38 (2007) 1485e1499.
[16] H. Nguyen-Thi, G. Reinhart, N. Mangelinck-Noël, H. Jung, B. Billia, T. Schenk,
J. Gastaldi, J. Härtwig, J. Baruchel, Metall. Mater. Trans. A 38 (2007)
1458e1464.
[17] C.A. Siqueira, N. Cheung, A. Garcia, J. Alloys Compounds 351 (2003) 126e134.
[18] J.E. Spinelli, I.L. Ferreira, A. Garcia, J. Alloys Compounds 384 (2004) 217e226.
[19] A. Badillo, C. Beckermann, Acta Mater. 54 (2006) 2015e2026.
[20] M.V. Canté, K.S. Cruz, J.E. Spinelli, N. Cheung, A. Garcia, Mater. Lett. 61 (2007)
2135e2138.
[21] H.B. Dong, X.L. Yang, P.D. Lee, W. Wang, J. Mater. Sci. 39 (2004) 7207e7212.
[22] J. Banaszek, S. McFadden, D.J. Browne, L. Sturz, G. Zimmermann, Metall. Mater.
Trans. A 38 (2007) 1476e1484.
[23] M. Wu, A. Ludwig, Metall. Mater. Trans. A 38A (2007) 1465e1475.
[24] V.B. Biscuola, M.A. Martorano, Metall. Mater. Trans. A 39 (2008) 2885e2895.
[25] A. Kumar, P. Dutta, J. Mater. Sci. 44 (2009) 3952e3961.
[26] J.N. Silva, D.J. Moutinho, A.L. Moreira, I.L. Ferreira, O.L. Rocha, J. Alloys Compd.
478 (2009) 358e366.
[27] A. Noeppel, O. Budenkova, G. Zimmermann, L. Sturz, N. Mangelinck-Noël,
H. Jung, H. Nguyen-Thi, B. Billia, C.A. Gandin, Y. Fautrelle, Int. J. Cast Met. Res.
22 (2009) 34e38.
[28] H. Jung, N. Mangelinck-Noël, H. Nguyen-Thi, B. Billia, J. Alloys Compd. 484
(2009) 739e746.
[29] S. McFadden, D.J. Browne, Appl. Math. Modell. 33 (2009) 1397e1416.
[30] M. Wu, A. Fjeld, A. Ludwig, Comp. Mater. Sci. 50 (2010) 32e42.
[31] M. Wu, A. Fjeld, A. Ludwig, Comp. Mater. Sci. 50 (2010) 43e58.
[32] A.E. Ares, S.F. Gueijman, C.E. Schvezov, J. Cryst Growth 312 (2010) 2154e2170.
[33] X. Li, Y. Fautrelle, K. Zaidat, A. Gagnoud, Z. Ren, R. Moreau, Y. Zhang, C. Esling,
J. Cryst. Growth 312 (2010) 267e272.
[34] H.B. Dong, P.D. Lee, Acta Mater. 53 (2005) 659e668.
[35] S.P. Wu, D.R. Liu, J.J. Guo, Y.Q. Su, H.Z. Fu, J. Alloys Compounds 441 (2007)
267e277.
[36] J.S. Wettlaufer, M.G. Worster, H.E. Huppert, J. Fluid Mech. 344 (1997) 291e316.
[37] X.G. Zhang, Corrosion and Electrochemistry of Zinc, Plenum Press, New York
and London, 1996.
[38] M. Sahoo, L.V. Whiting, D.W.G. White, AFS Transactions 93 (1985) 475e480.
[39] B.K. Prasad, A.H. Padwardhan, Y. Yegneswaran, Z. fur Metallkunde 88 (1997)
333e338.
[40] P. Sriram, S. Seshan, H. Md, Trans. Am. Soc. 109 (1992) 769e775.
[41] W.R. Osório, C.M. Freire, A. Garcia, Mater. Sci. Eng. A 402 (2005) 22e32.
[42] W.R. Osório, C.M. Freire, A. Garcia, J. Alloys Compounds 397 (2005) 179e191.
[43] W.R. Osório, M.E.P. Souza, C.M. Freire, A. Garcia, J. New Mat. Electrochem.
Systems 11 (2008) 37e42.
[44] M.T. Abou El-khair, A. Danoud, A. Ismail, Mater. Lett. 58 (2004) 1754e1760.
[45] R.A. Auras, C.E. Schvezov, Metallurgical and Materials Transactions A 35
(2004) 1579e1589.
[46] I. Manna, J.N. Jha, K. Pabi, Scripta Metallurgica et Materialia 29 (1993) 817e822.
[47] I. Manna, S.K. Pabi, W. Gust, Int. Mater. Rev. 46 (2001) 53e91.
[48] J. Dutta Majumdar, A. Weisheit, B.L. Mordike, I. Manna, Mater. Sci. Eng. A 266
(1999) 123e134.
[49] J. Dutta Majumdar, R. Galun, B.L. Mordike, I. Manna, Mater. Sci. Eng. A 361
(2003) 119e129.
[50] I. Manna, S.K. Pabi, W. Gust, Acta Metall. Mater. 39 (1991) 1489e1496.
[51] D.B. Williams, E.P. Butler, Inter. Met. Rev. 26 (1981) 153e183.
[52] W. Gust, in Phase Transformations, Series 3, N 11, vol. 1, The Institute of
Metallurgists (ed.), p. II/27, The Chameleon Press, London (1979).
[53] M. Friesel, I. Manna, W. Gust (Colloque), J. de Physique 51 (1990). C1-381-C1-390.
[54] I. Manna, W. Gust, B. Predel, Scripta Metall. Mater. 24 (1990) 1635e1640.
[55] B. Predel, W. Gust, Mater. Sci. Eng. 16 (1974) 239e249.
[56] S.P. Gupta, Acta Metall. 35 (1987) 747e757.
[57] C.P. Ju, R.A. Fournelle, Acta Metall. 33 (1985) 71e81.
[58] I. Manna, J.N. Jha, S.K. Pabi, W. Gust, in: Structure and properties of interfaces
in materials, materials research Society Symp. Proc., W.AT. Clark, U. Dahmen &
C.L. Briant (eds.), Materials Research Society, Pittburgh, USA 238 (1992) 517.
[59] S.F. Gueijman, C.E. Schvezov, A.E. Ares, Mater. Trans. 51 (2010) 1861e1870.
[60] A.E. Ares, I.P. Gatti, S.F. Gueijman, C.E. Schvezov, Proceedings of the Modeling
of Casting, Welding and Advanced Solidification Processes e XII, The Minerals,
Metals and Materials Society, Rosewood Drive, Danvers, USA, 2009.
[61] A. E. Ares, L. M. Gassa, S. F. Gueijman, C. E. Schvezov, Proceedings of shape
casting: the 3rd International Symposium, The Minerals, Metals and Materials
Society, Rosewood Drive, Danvers, MA 01923, USA, 2009.
[62] M.R. Rosenberger, Alicia E. Ares, Isaura P. . Gatti, Carlos E. Schvezov, Wear 268
(2010) 1533e1536.
[63] A.E. Ares, L.M. Gassa, S.F. Gueijman, C.E. Schvezov, J. of Crystal Growth 310
(2008) 1355e1361.
[64] A.E. Ares, L.M. Gassa, Corros. Sci. 59 (2012) 290e306.
[65] R.J. Barnhurst, Zinc and zinc alloy casting, in: R.J. Barnhurst (Ed.), ASM Hand
Book, tenth ed., vol. 2, ASM International, The Materials Information Society,
Metals Park, OH, 1995.
[66] R.J. Barnhurst, E. Gerevais, AFS Transaction 85-91 (1983) 591e602.
[67] N. Tunca, R.W. Smith, J. Mat Sci. 23 (1988) 111e120.
[68] R. Trivedi, W. Kurz, Int. Mater. Rev. 39 (1994) 49e74.
[69] H.E. Boyer, T.L. Gall, Metals Handbook. Desk Edition, American Society for
Metals, USA, 1990.
[70] M.R. Rosenberger, C.E. Schvezov, E. Forlerer, Wear 259 (2005) 590e601.
[71] M. Vandyoussefi, A.L. Greer, Acta Materialia 50 (2002) 1693e1705.
[72] Ch. A. Gandin, ISIJ Int. 40 (2000) 971e979.
[73] W.R. Osório, A. Garcia, Mater. Sci. Eng. A 325 (2002) 103e111.
[74] J. Augustynski, Corros. Sci. 13 (1978) 955e965.
[75] H. Morrow, in: M.B. Bever (Ed.), Encyclopedia of Materials Science and Engi-
neering, vol. 7, MIT Press, Cambridge, Massachusetts, 1986.
[76] I. Epelboin, M. Keddam, H. Takenouti, J. Appl. Electrochem. 2 (1972) 71e79.
[77] L. Fedrizzi, L. Ciaghi, P.L. Bonora, R. Fratesi, G. Roventi, J. Appl. Electrochem. 22
(1992) 247e254.
[78] D.A. Rigney, Wear 245 (2000) 1e9.
[79] B. Venkataraman, G. Sundararajan, Acta Mater. 44 (1996) 461e473.